Stereotactic Body Radiation Therapy Planning For Localized Prostate Cancer In The Pace Phase 3 Trial: Can This Be Done Without Fused Planning Mri?

The PACE trial randomizes between stereotactic body radiation therapy (SBRT) and either prostatectomy or conventionally fractionated radiation therapy for localized prostate cancer (NCT01584258; UKCRN12628). A fused magnetic resonance imaging (MRI) planning scan has been mandated in PACE, but this may limit recruitment at centers where it is not routine. Studies have suggested computed tomography (CT)-only volumes may increase dose to the rectum. We aimed to determine whether CT-only volumes would be similar enough to MRI-fusion volumes to produce acceptable PACE SBRT plans. CT data sets from 20 patients previously treated according to the PACE trial protocol were used. These patients had fused MRI-aided contouring (MR). Treatment was 36.25 Gy in 5 fractions to the planning target volume (PTV) (with prostate itself receiving 40 Gy in 5 fractions). Patients were anonymously recontoured on CT only (with no reference to MRI data) on 2 separate occasions, 2 months apart (CT1 and CT2). Contouring was done as a consensus volume of 2 observers experienced in both CT-only and MR-fusion contouring. These patients were then replanned by an independent planner with CT-only volumes and no reference to the original plan. Contouring and planning was done as per PACE protocol, using a robotic radiosurgery inverse planning system. Dice coefficient was used to assess volume concordance. Mean prostate and base of seminal vesicle volumes were 63.5 mL (range 23-144), 63.2 (range 33-143), and 63.8 (range 32-142) for MR, CT1, and CT2, respectively. Mean Dice coefficient was 0.86 (CI 0.84-0.87), 0.85 (0.83-0.86), 0.92 (0.91-0.93) for MRvCT1, MRvCT2, and CT1vCT2, respectively. The apex was on average contoured 1.1 mm lower on MR than with CT (SD 3.5 mm). Eighteen of the MR-contoured plans met all PACE dose constraints, versus 19 of the CT-contoured plans. Median rectal V18.1Gy, V29Gy and V36Gy and bladder V18.1Gy, V37Gy were not significantly different between MR- or CT-contoured plans (Wilcoxon signed ranks test). CT-only contouring produces similar volumes to fused MRI contours and no dosimetric disadvantage. While we would still strongly recommend MRI-fusion contouring for its superior anatomical fidelity, the PACE protocol has been modified such that this is no longer mandated.